Chemistry Hougang: Chemical Analysis Explained

“How do you know?” is one of the most important questions in Chemistry. A bottle can be clear and still contain dissolved substances. A solution can change colour without identifying one unique chemical. A machine can show a long row of numbers while the sample was collected incorrectly. The art of analytical Chemistry is not simply getting a signal. It is deciding what that signal can legitimately tell us.

This Chemistry Hougang guide is for families searching for help with Chemistry experiments, qualitative analysis, spectroscopy, JC Chemistry practical skills and reliable scientific evidence. A Chemistry tutor can teach students to name tests and copy results, but I want the learner to grow beyond that: identify the question, design the observation, understand the measurement, check uncertainty, and explain what conclusion is justified.

The wider view is remarkable. A student learning to read a burette in Secondary school is beginning the same intellectual journey as a university analyst checking a medicine’s purity, a scientist measuring substances in river water, or a geochemist studying a mineral sample. The equipment becomes more sophisticated. The obligation to tell the truth about the evidence stays the same.

What is chemical analysis?

Chemical analysis is the use of appropriate methods to learn about a material or sample, often including which substances are present and how much of them there is. A qualitative question asks about identity or presence: what might be in the sample? A quantitative question asks about amount or concentration: how much is present under defined conditions?

The answers can be related but require different evidence. Detecting a signal associated with a substance does not automatically give an accurate concentration. Reporting a concentration is unreliable if the measured signal is not specific to the intended substance.

Real-world analytical work must also account for sampling and preparation. An accurate measurement of an unrepresentative sample can still lead to a misleading conclusion about a whole tank, field, batch or river. A sample may change through poor storage, contamination or loss before it reaches the instrument.

I teach the process as seven connected questions: what are we asking, what did we collect, what method fits, how was it calibrated, how was it checked, what does the result mean, and what remains uncertain?

Start with the question, not the instrument

Imagine a parent asking whether water from a source is safe. An inexperienced response may begin with a pH test. A pH measurement can be relevant, but it cannot by itself assess microbial hazards, harmful concentrations of metals, pesticides or other contaminants. Safe drinking water requires a much broader, context-dependent evaluation.

Now change the task. The question is whether two samples have different pH values at the time of measurement. A properly maintained and calibrated pH meter may be very useful. The *same instrument* that was insufficient for the first question can be appropriate for the second.

This is the key lesson: a result is only as useful as its fitness for the question being asked. A student should never equate having an instrument with having enough evidence.

We can apply this to food, medicines, environmental monitoring, materials, art conservation and forensic investigations. What is the analyte? In what material or matrix is it found? What other substances might interfere? How precise does the answer need to be for the decision?

Sampling is part of the chemistry

Suppose one spoonful is taken from a poorly mixed container, and the analyst measures it with exceptional precision. If the container’s composition varies, that spoonful may not represent the whole. Greater decimal precision does not repair bad sampling.

Chemists consider where and when samples were collected, how containers were prepared, whether the target substance is stable, and how potential contamination was controlled. For a changing river, a measurement at one location and time is not automatically a complete description of an entire waterway.

I find this the first big conceptual leap for students. In a school question, the sample may be treated as if it is already valid. In real scientific work, the preparation and selection of the sample are themselves hypotheses about what the collected material represents.

A clear report therefore distinguishes sample concentration, site or batch inference, and the limits of the sampling design. They are different levels of claim.

What is the difference between accuracy, precision and reliability?

These words are related but should not be collapsed into one meaning.

Precision describes how closely repeat measurements agree under specified conditions. Accuracy concerns closeness to the true or accepted reference value, although metrology defines related terms with care. Uncertainty characterises the range or dispersion of values that can reasonably be attributed to what is being measured under a stated method.

Imagine an instrument repeatedly reporting 5.00, 5.01 and 4.99 units for a reference material whose accepted value is 6.00 units. The numbers agree closely, so the readings show high repeatability. Yet they may be systematically biased, so the reported result is not close to the reference value.

That is why repeating the same flawed measurement ten more times is not a substitute for examining calibration, method bias or the reference.

The National Institute of Standards and Technology explains measurement traceability, which involves a documented chain of calibration relationships and associated uncertainties. A professional measurement needs more than a machine labelled “precise”.

The invisible ingredients of a trustworthy test

I ask students to learn five laboratory words even before they are expected to use advanced instruments.

Blank: a relevant material treated through the method without the target analyte, used to detect background response or contamination.
Standard: a material or solution with a known quantity used to check or establish an instrument’s response.
Control: a material or procedure used to check that a method is behaving as intended.
Replicate: a repeated measurement or sample preparation used to assess variability at a specified stage.
Matrix: the wider sample environment surrounding the substance being measured, which can influence the analytical result.

Each word answers a different question. A blank does not automatically prove there is no matrix interference. Replicates do not automatically prove an instrument has no bias. A calibration standard does not guarantee that an unknown sample behaves exactly like the standards.

This is what makes analytical Chemistry intellectually interesting: every useful check reduces one kind of doubt, but different doubts require different tests.

Worked example: a calibration curve that includes a background signal

Consider a completely fictional laboratory exercise. An instrument measures a signal related to the concentration of substance X in a solution. Four prepared calibration standards yield the following idealised values:

0 mg/L gives a signal of 0.10.
2 mg/L gives a signal of 0.30.
4 mg/L gives a signal of 0.50.
6 mg/L gives a signal of 0.70.

In this artificial example the relationship is linear over the stated range. Each increase of 2 mg/L increases the signal by 0.20, giving a slope of 0.10 signal units per mg/L. At concentration zero, the signal is still 0.10, so the calibration intercept matters.

The model is signal = 0.10 + (0.10 × concentration in mg/L).

An unknown prepared sample gives a signal of 0.45. We solve:

0.45 = 0.10 + 0.10c, so c = (0.45 − 0.10)/0.10 = 3.5 mg/L.

Notice what would happen if the student wrongly forced the calibration through the origin. Their concentration estimate would change. The error would not be a calculator error; it would be a modelling error about background signal.

The NIST guide to calibration in chemical measurement explains why choosing and evaluating an appropriate calibration relationship matters. Real calibrations require standards, uncertainty, checks on model suitability and attention to the sample matrix. Our example is a mathematical teaching model, not a validated analytical method.

What if the unknown sample had been diluted?

Suppose the technician took one volume of original sample and diluted it to four volumes in total, then measured that diluted sample. The instrument’s calculated 3.5 mg/L refers to the diluted solution, assuming the method was otherwise suitable.

With a total-volume dilution factor of 4, the corresponding original-sample concentration would be 3.5 × 4 = 14 mg/L. The units must stay consistent. Merely adding four volumes of water *to* one volume of sample would instead make the total volume five volumes and require a different factor.

This difference is a classic examination trap. A student may know the formula C₁V₁ = C₂V₂, but if they misunderstand “diluted to” versus “mixed with”, the mathematics cannot rescue their interpretation.

A careful report states the preparation, dilution factor, measured result, calculated original concentration and uncertainty. It should not quietly jump from an instrument’s signal to a confident claim about the whole original material.

Two repeated results are not automatically enough

Imagine the diluted fictional sample is measured twice, giving 3.4 and 3.6 mg/L. The mean is 3.5 mg/L, and the range is 0.2 mg/L. A common simple measure is relative percent difference, calculated here as:

|3.6 − 3.4| / 3.5 × 100% ≈ 5.7%.

That calculation quantifies how far the two results differ relative to their mean. It does not, by itself, establish that the method meets an appropriate quality standard. Acceptance criteria depend on the analyte, method, concentration, uncertainty and the decision being made. Two closely agreeing results could both be biased.

This is an example of a numerical answer that becomes useful only when it sits inside a justified quality-control framework.

How classical chemical tests differ from instrument-based analysis

School Chemistry often uses colour changes, precipitates, gases, flame observations and related qualitative tests. These can teach important patterns and decision processes. But I encourage students not to overstate specificity.

A single visible observation may be consistent with more than one cause, particularly when interfering substances or different reaction conditions are possible. Identifying an unknown often requires a sequence of observations with appropriate controls, rather than one guess based on colour.

Professional methods may use chromatography, spectroscopy, mass spectrometry and electrochemical methods. Each has characteristic strengths and limitations.

Chromatography separates components under the chosen method according to how they interact with a stationary and mobile phase. A retention time can aid identification but does not necessarily identify a substance uniquely on its own.

Spectroscopy examines interactions between matter and electromagnetic radiation. Depending on the technique, it can provide information about concentration, functional groups, elements, electronic transitions or molecular structure.

Mass spectrometry examines ions according to mass-to-charge ratio and related fragmentation or ionisation information. It can provide powerful evidence about chemical identity, especially when combined with separation methods, reference standards and appropriate interpretation.

The NIST discussion of quantitation in chromatography reminds us that sample handling, standards, replication, instrumental drift and uncertainty are part of getting meaningful numbers.

A spectrum is evidence, not a magic fingerprint

A novice might say that one peak in a spectrum proves the identity of an entire sample. A trained analyst asks what instrument produced the spectrum, whether the signal is specific, what could overlap, whether the sample is a mixture, and what reference or complementary evidence supports the conclusion.

The exact logic differs among infrared spectroscopy, UV–visible spectrophotometry, elemental spectroscopy and mass spectrometry. The school-level takeaway is that an instrument output is a representation of a measurement, not a direct photograph of molecules.

Students who have learned to move from an observation to a particle explanation and a symbolic model are well prepared for this. The more advanced instrument simply increases the need to control the assumptions.

We explore this way of thinking in How to Think Like a Chemist, and a deeper example in The Gold Standard of Spectroscopy.

The problem with ‘not detected’

A laboratory report saying “not detected” does not automatically mean that absolutely none of the substance is present. It often means that the analytical response did not meet the relevant detection or reporting decision criterion under a particular method.

The difference between a detection limit and a quantitation limit can be important. Detection concerns whether a signal can be distinguished from an appropriate background, while quantitation concerns reporting an amount with adequate method performance at the relevant level. The exact definitions and rules depend on the applicable analytical framework.

The US EPA’s laboratory QA/QC explanation describes how detection limits, quality controls and reported measurements depend on defined analytical methods.

A scientific reader must resist two opposite errors: claiming “zero” where the method could only establish “below the reporting level”, and claiming “danger” from a detected amount without understanding concentration, exposure and the relevant decision criteria.

When a number becomes a public decision

Consider a city trying to understand a possible contaminant in drinking water, a food manufacturer checking ingredients or a healthcare laboratory interpreting a result. These decisions have important consequences. The ethical responsibility is to connect the analytical number to the right question and the right standard, while documenting uncertainty.

A good laboratory report separates measured results, method descriptions, interpretation and limits. It should identify the sampling context, reporting units and any relevant qualification. It should not hide a failed quality-control check to make a conclusion look more impressive.

There is also a societal dimension. If a test is expensive, which communities have access to it? If sample collection is uneven, whose environment is represented? If a company’s commercial interests conflict with transparent reporting, what procedures protect the integrity of the evidence? Chemists work with statistical, engineering, governance and ethical partners to answer these questions.

This is why chemical analysis belongs within education as a way of living responsibly with evidence, not just as an examination chapter.

What Secondary students should master first

For a Secondary 1–2 student, I would prioritise careful reading of variables, sensible units, tables and graphs, and the difference between observation and inference. Students must learn to write accurately rather than use an unsupported term because it sounds scientific.

For Secondary 3–4 Chemistry, I would add correct ionic tests to the appropriate syllabus, balanced equations where relevant, clear practical descriptions, proportional reasoning, concentrations, significant figures and error analysis. They should be able to explain why a particular test or observation supports a conclusion and which other results are needed.

At JC H1 or H2 level, quantitative and practical questions demand more detailed uncertainty reasoning and connections to equilibria, reaction conditions and analytical methods. Students who know every calculation but have never checked the meaning of a measured volume are vulnerable in complex questions.

The important shift is from reproducing an experiment procedure to understanding what evidence the procedure is designed to produce.

What university and working analysts add

University laboratory work may require method development, calibration models, validated procedures, reference materials, reporting of uncertainty, instrumental troubleshooting, sample traceability, maintenance and repeatable documentation.

In a research setting, the question may be novel, so no standard procedure gives the answer automatically. Chemists must design studies, establish controls and weigh alternative interpretations. In regulated work, compliance and documentation are part of quality. Laboratory decisions are often reviewed by other specialists; clear communication is as important as technical fluency.

I like that there is a continuous line from the student’s first correctly labelled graph to a defensible professional analytical report. The advanced scientist should still be prepared to answer the same basic question: How do we know?

A four-week study plan that builds genuine evidence literacy

No home chemistry set or chemical mixing is needed. The exercises below use reading, paper problems and publicly available information.

Week 1 — Define the question. Compare two claims: “the water has pH 7” and “the water is safe to drink”. Explain why the first cannot by itself establish the second. Name two further kinds of evidence that could be relevant.

Week 2 — Understand calibration. Reproduce the fictional four-point calibration exercise above. Change the unknown signal to 0.55 and find the new diluted concentration: (0.55 − 0.10)/0.10 = 4.5 mg/L, under the same idealised model.

Week 3 — Trace the sample. Draw a flow chart from sample collection through preparation, measurement, controls, calculation and final report. Write one possible error or uncertainty source at each step.

Week 4 — Explain a result. Produce a 250-word report about the fictional sample. State the analyte, sample context, calibration result, dilution factor, limitations and what cannot be concluded about safety or legal compliance.

I would assess the final report for honest qualification, correct units and causal reasoning. Those habits transfer from Chemistry to Physics, Biology, Geography, medicine and many forms of professional work.

A diagnostic test with full answers

Question 1: the colour-change shortcut

An unknown solution changes colour in a general reaction. Is its identity fully confirmed?

Answer: not necessarily. The observation may be consistent with several substances or with interference. An appropriate confirmatory method and controls may be required.

Question 2: zero concentration versus zero signal

A calibration standard at 0 mg/L gives a signal of 0.10. Can the analyst assume all unknown signals start at zero?

Answer: no. A nonzero background or intercept must be addressed by a suitable calibration model and appropriate blank checks.

Question 3: calculating with the correct intercept

For signal = 0.10 + 0.10c, an unknown gives signal 0.60. Find c.

Answer: c = (0.60 − 0.10) / 0.10 = 5.0 mg/L, assuming the calibration is valid at that response and concentration.

Question 4: the dilution trap

A technician puts one part sample into a container and adds four more equal-volume parts of diluent. What is the total-volume dilution factor?

Answer: 5, not 4, assuming volumes are additive as stated. The total final volume is five parts.

Question 5: replicate readings

Two readings agree almost perfectly. Is that enough to establish they are accurate?

Answer: no. Agreement indicates repeatability under the measurement conditions but cannot eliminate shared calibration bias, sampling error or matrix interference.

Question 6: ‘not detected’

Does a “not detected” result guarantee an analyte is absent?

Answer: no. It usually refers to the method’s defined ability to detect or report a signal under the stated conditions. The method and limit are essential for interpretation.

Parents: what makes practical Chemistry tuition effective?

When a student loses marks in practical or data questions, I examine the precise failure. Did they misread the command word? Did they confuse units? Did they treat an inference as a direct observation? Did they calculate with the initial rather than final volume? Did they assume a test was unique?

A useful Chemistry tutor supporting students in Hougang should diagnose these separately. A tutor who simply hands the learner another set of practical papers may reinforce the same habits without repairing the misunderstanding.

I would use one fully explained example, one near-transfer example and one unfamiliar case. The student needs to state both the conclusion and its limit. In time, they should begin to find their own errors rather than wait for a teacher to point them out.

A parent’s best question after the lesson is simple: “Show me something you originally concluded too confidently, and how the evidence changed your answer.” That demonstrates authentic scientific growth.

Frequently asked questions

Is qualitative analysis just memorising colours and gas tests?

No. Knowing prescribed observations is necessary, but a strong answer also connects tests to the correct chemical principles, conditions and possible alternatives, as appropriate to the syllabus.

Why are significant figures important in Chemistry?

They help prevent reporting implausible precision, but they do not replace a complete treatment of measurement uncertainty. Correct units, instrument resolution, calibration and method assumptions also matter.

Do professional chemists still use titration?

Yes, volumetric methods remain useful for suitable problems. Whether titration or an instrumental technique is appropriate depends on the analyte, matrix, required performance and available resources.

Is one advanced instrument enough to identify every substance?

No. All methods have scopes and limitations. Complementary methods, reference standards and suitable controls may be required.

How does analytical Chemistry connect to university and work?

It underpins environmental testing, pharmaceutical quality, food science, forensic applications, materials research, manufacturing, public health and scientific research, with different training and regulatory requirements.

Sources and further reading

Keep exploring Chemistry and evidence

If a student is struggling, their latest experiment write-up, practical calculation and written explanation will tell us much more than their confidence alone. Parents who want a learning diagnosis can arrange a consultation with eduKate Singapore.

Chemistry Hougang | Chemical Analysis Explained: Experiments, Spectroscopy and Reliable Evidence